Catch Roles Early: Alerts, Search Operators and Public Job Feeds

Two students apply for the same junior role. One applies on the second day after it was posted, when the recruiter has twenty applications to look through. The other applies on the tenth day, when there are four hundred. Nothing else is different, but the first student's resume gets a real look and the second's competes for seconds of attention.

You cannot check fifty careers pages every morning. You can make new roles come to you. This article sets that up, from simple alerts anyone can create to a small script for those who code.

• Why applying early matters, and what "early" realistically means.
• How to set up narrow alerts on job boards and LinkedIn.
• Boolean search and search-engine site: queries that surface roles directly on company job boards.
• How to watch careers pages and public job feeds for changes.
• An optional, responsible automation that filters new roles and messages you.

Recruiters often review applications as they arrive, and attention is highest in the first days a role is open, before the queue grows and before they may have enough strong candidates to start interviews. There is no magic hour, and claims about precise windows are guesswork. But applying within a few days of a role being posted is a simple advantage, and it is mostly a matter of finding out quickly.

Companies title the same entry-level role many ways: "Software Engineer I", "Associate Software Engineer", "SDE-1", "Backend Developer", "Graduate Engineer", "Java Developer". One broad alert ("software engineer Bengaluru") returns hundreds of irrelevant results, while missing some of these titles. Several narrow alerts work better:

Alert 1: "backend developer" Java, 0 to 2 years, BengaluruAlert 2: "associate software engineer", 0 to 1 years, Bengaluru or remoteAlert 3: "SDE 1" OR "software engineer I", BengaluruAlert 4: "Spring Boot developer", fresher, Bengaluru or Pune

Set them on each job board you use and on LinkedIn Jobs, with "past 24 hours" or "past week" where the site offers it, and daily email or app notifications.

LinkedIn's job search and several job boards accept Boolean operators in the keyword box. They let you include alternatives and exclude noise.

A useful fresher query:

("software engineer" OR "backend developer" OR "SDE") AND (Java OR Python OR Node)NOT (senior OR lead OR manager OR principal OR staff)

Exact support varies by site and changes over time, so test a query and compare the results with and without each part.

Many companies host their job listings on applicant tracking systems with public job-board pages, at addresses such as boards.greenhouse.io/<company>, job-boards.greenhouse.io/<company> or jobs.lever.co/<company>. Search engines index many of these pages, so a site: query finds roles across many companies at once, often before job boards repost them:

site:boards.greenhouse.io OR site:job-boards.greenhouse.io "software engineer" Indiasite:jobs.lever.co ("associate" OR "graduate" OR "new grad") engineer Bengaluru

Add a time filter in the search engine's tools (past week) to see recent pages. Search engines index these pages with a delay, so treat this as discovery, not real time. Then open the company's own careers page to confirm the role is open.

For your top 15 to 20 companies, you want to know as soon as a new role appears. Three ways, from simplest:

1. The site's own alerts. Some careers pages offer "get notified" or job alert emails. Use them.
2. Page-change monitoring. Tools such as the open-source changedetection.io (which you can run yourself or use as a hosted service) or Visualping check a page on a schedule and notify you when it changes. Point them at the careers page filtered to engineering, or better, at the company's job-board page, which changes less for cosmetic reasons.
3. Public job feeds (next section), which change only when roles are added or removed.

Free plans of hosted monitoring tools limit how many pages and how often; check the current limits.

Some applicant tracking systems publish public, read-only job-board APIs so that companies can show their roles on their own websites. Anyone can read them, without an account, and they return structured data (JSON) about every open role. Two widely used examples, as of October 2026:

Greenhouse job board API:  https://boards-api.greenhouse.io/v1/boards/<board-token>/jobsLever postings API:        https://api.lever.co/v0/postings/<company>?mode=json

The board token or company name is the part after the domain in the company's job-board address (for example, boards.greenhouse.io/examplecorp gives examplecorp). Opening these addresses in a browser shows the raw data. A feed reader cannot read JSON directly, but a small script can.

This program filters a feed for fresher-friendly roles in India or remote, and reports only roles it has not seen before. To run offline, it uses a made-up sample shaped like a job-board API response. A real version would download the feed and store the seen IDs in a file between runs.

# File: job_feed_filter.py# Filter a job-board feed for fresher-friendly roles and report new ones.# SAMPLE is shaped like the JSON a public job board API returns. It is made-up# data so the program runs offline; a real script would download the feed.import jsonSAMPLE = """{"jobs": [  {"id": 101, "title": "Software Engineer I, Backend", "location": {"name": "Bengaluru, India"},   "absolute_url": "https://example.com/jobs/101"},  {"id": 102, "title": "Senior Software Engineer, Payments", "location": {"name": "Bengaluru, India"},   "absolute_url": "https://example.com/jobs/102"},  {"id": 103, "title": "Associate Software Engineer", "location": {"name": "Remote, India"},   "absolute_url": "https://example.com/jobs/103"},  {"id": 104, "title": "Product Designer", "location": {"name": "Bengaluru, India"},   "absolute_url": "https://example.com/jobs/104"},  {"id": 105, "title": "Backend Engineer (0-2 years)", "location": {"name": "Pune, India"},   "absolute_url": "https://example.com/jobs/105"},  {"id": 106, "title": "Software Engineer, Graduate", "location": {"name": "London, UK"},   "absolute_url": "https://example.com/jobs/106"}]}"""WANT = ("software engineer", "backend", "developer", "sde")AVOID = ("senior", "staff", "lead", "principal", "manager")PLACES = ("india", "remote")ALREADY_SEEN = {101}  # ids from earlier runs; a real script would store these in a filedef is_match(job):    title = job["title"].lower()    place = job["location"]["name"].lower()    return (any(w in title for w in WANT)            and not any(a in title for a in AVOID)            and any(p in place for p in PLACES))jobs = json.loads(SAMPLE)["jobs"]matches = [j for j in jobs if is_match(j)]new = [j for j in matches if j["id"] not in ALREADY_SEEN]print(f"{len(jobs)} roles in feed, {len(matches)} match, {len(new)} new")for job in new:    print(f"NEW  {job['title']} | {job['location']['name']} | {job['absolute_url']}")

6 roles in feed, 3 match, 2 newNEW  Associate Software Engineer | Remote, India | https://example.com/jobs/103NEW  Backend Engineer (0-2 years) | Pune, India | https://example.com/jobs/105

Read the output against the sample: the senior role and the designer role were filtered out by title, the London role by location, and role 101 matched but was already seen. Simple keyword rules make mistakes in both directions (a "Lead Generation Engineer" would be wrongly excluded), which is why you still read the shortlist yourself.

If you enjoy building things, you can combine these pieces with a workflow tool such as n8n (open source, can be self-hosted) or a similar service with a free tier, or a scheduled script:

Every few hours:  fetch the public feeds of your target companies  keep roles not seen before  filter by title, seniority and location  (optional) ask an AI model to rate fit against a short profile  send the shortlist to yourself (email or a messaging app)  add each role to your tracker with status "to review"

Three cautions for the AI step:

• It is a filter, not a judge. An AI rating can misread a description. Use it to sort, then read every role it rates highly, and glance at the ones it rated low.
• Do not send personal data to the model. A short profile ("fresher, Java and Spring Boot, Bengaluru or remote") is enough.
• Mind costs and limits. Free API tiers have limits that change. A keyword filter alone may be all you need.

Building this pipeline is itself a decent small project for your GitHub, as long as it respects every site's terms.

Nikhil has 18 high-priority companies. He checks how each lists jobs:

• 9 use job-board pages with public feeds. He writes a 60-line script based on the filter above, runs it twice a day, and gets an email listing new matches.
• 5 have their own careers pages without feeds. He sets up page-change monitoring on each company's engineering jobs page.
• 4 post mainly on job boards and LinkedIn. He sets narrow alerts for their names plus "engineer".

In the next three weeks, his tracker shows he applied to 11 roles within two days of posting, against 3 in the three weeks before.

• One broad alert that floods your inbox and misses alternative titles.
• Trusting a site: search as real time. Search engines index with a delay; confirm on the careers page.
• Hammering sites with frequent requests, or scraping sites that forbid it.
• Letting an AI filter decide without reading the shortlist.
• Building automation instead of applying. A spreadsheet of alerts and an hour a day of applications beats a beautiful pipeline that took your whole week.

Alert 1: "data analyst", 0 to 2 years, Hyderabad or remoteAlert 2: "business analyst" AND (SQL OR "Power BI"), fresher, HyderabadAlert 3: "MIS analyst" OR "reporting analyst", 0 to 1 years, HyderabadAlert 4: ("analytics associate" OR "junior analyst" OR "associate analyst") AND SQL, remote India

("software engineer" OR "software developer" OR "SDE") AND (Java OR Python OR React)NOT (senior OR lead OR manager)

• Next: Find Companies Before They Post: Funding News and Founder Signals finds companies likely to hire soon.
• For AI workflows and automation done reliably, see the AI course's article Where Agents Fail and How to Build Reliable AI Workflows.
• For the full job search plan around these alerts, see How to Get Your First Job as a Fresher in India: A Step-by-Step Guide.